{"id":"456c45c3-9bf8-440d-a4f1-46b4018948c7","arxiv_id":"1908.08720","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Ångström-resolved Raman images of individual vibrational modes of a single Mg-porphine molecule are used to demonstrate a Lego-like assembly protocol for determining chemical structure.","lead":"Scientists used an ultra-sharp silver tip and laser light to take Raman pictures of a single molecule with a resolution of about 1.5 angstroms, smaller than a chemical bond. They propose a Lego-like method, called scanning Raman picoscopy, to rebuild a molecule's chemical structure from these images.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 1.5 Å resolution claim is under-sampled: 25×25 pixels over 2.5 nm gives ~1.0 Å pitch, below Nyquist for a 1.5 Å FWHM, so bond-resolving images and the Lego structure may rest on interpolation artifacts.","rationale":"The reader's conditional verdict rests on the plausibility of Ångström-resolved TERS images and the deferred supplementary data. My concern identifies a concrete, checkable problem in the main text's evidence for the headline resolution: the stated pixel pitch is insufficient to support a 1.5 Å FWHM. This does not prove the method is wrong, but it shifts the burden: the strongest version of the claim—bond-resolving structural determination—is not established by the images as described. Because the supplementary material might contain higher-density scans or raw data, the appropriate verdict remains conditional rather than outright rejection; the test above would decide whether the concern lands. I partially agree with the reader's weakest-assumption identification: the localized-field width is indeed load-bearing, and pixelation is the most direct way to test whether the field is actually that narrow.","tokens_in":9366,"tokens_out":10158,"duration_ms":110938,"concrete_test":"Take the raw 25×25 pixel data for the 3072 cm−1 mode and recompute the line-profile FWHM without interpolation, counting the number of non-interpolated pixels above half maximum. If fewer than three independent pixels fall within the FWHM, or if fitting with linear, spline, and sinc interpolation changes the FWHM by more than 30%, the 1.5(1) Å resolution claim is not supported. As a stronger test, re-acquire the same C–H mode over the same area with 50×50 or 100×100 pixels (or a smaller scan window at fixed 25×25); if the eight-dot pattern and the dark inter-dot minima change, the bond-resolved interpretation of Fig. 2c is an artifact of sampling.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The manuscript's central resolution claim—1.5(1) Å FWHM for the 3072 cm−1 mode (Fig. 1e)—is derived from a 25×25 pixel image spanning 2.5 nm (Fig. 1c caption). The pixel pitch is therefore ~1.0 Å, coarser than half the claimed FWHM. A 1.5 Å feature sampled at 1 Å intervals has at most two or three independent samples above half maximum, so the quoted FWHM and its 0.1 Å uncertainty cannot be robustly determined without raw unprocessed data or explicit interpolation assumptions. This matters because the single-chemical-bond resolution is the foundation for interpreting the eight bright dots in Fig. 2c as eight C–H bonds and the dark gaps as destructive interference between neighboring bonds. If the true optical spot is 2–3 Å wide, the bright/dark pattern would not map one-to-one onto bond positions, and the Lego construction in Figs. 2–4 would lose its experimental basis. The referenced computed SRP images in Supplementary S4 are useful independent support, but they cannot validate the experimental FWHM if the underlying scan is under-sampled. The concern is not about the existence of sub-nm TERS contrast, which is supported by prior work, but specifically about the quantitative Ångström resolution and the bond-level interpretation built on it.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports scanning Raman picoscopy (SRP), a tip-enhanced Raman spectroscopy (TERS) method in which a Raman spectrum is acquired at each pixel of a 25 × 25 scan over a single Mg-porphine molecule adsorbed on Ag(100). The authors claim Ångström-level spatial resolution, with a 1.5(1) Å FWHM line profile for the 3072 cm−1 mode, and present spatial maps for a number of vibrational modes that show distinct patterns. They interpret the images through an interference effect between neighboring bonds or rings, encoded in the sign of a cross-term in Eq. (2), and use these patterns together with Raman group frequencies to assemble the molecular structure in a 'Lego-like' fashion. The final structure is reported as fully determined in real space, and computed SRP images (Supplementary S4) are said to agree with experiment, validating the proposed methodology.","tokens_in":9619,"tokens_out":4156,"duration_ms":45957,"significance":"If the claims hold, SRP would be a significant advance: it would extend single-molecule TERS from spectral fingerprinting to bond-resolved real-space imaging and offer a route to structural determination of surface-adsorbed molecules. The experimental data are rich and the reported mode-specific images are visually compelling. The authors also make a genuine methodological proposal that combines Raman fingerprints, spatial mode images, and an interference rule to constrain molecular connectivity. However, the quantitative resolution claim is not supported by the stated sampling, and the structural reconstruction relies heavily on prior knowledge of the molecule and on an asserted sign rule; these issues must be addressed before the central claims can be accepted.","major_comments":[{"comment":"The claimed 1.5(1) Å spatial resolution is under-sampled relative to the stated acquisition parameters. The SRP image in Fig. 1c is acquired over 2.5 nm × 2.5 nm with 25 × 25 pixels, giving a 1.0 Å pixel pitch. Nyquist sampling of a 1.5 Å FWHM feature requires a pitch of about 0.75 Å or smaller, so the line profile in Fig. 1e cannot robustly determine the FWHM or its 0.1 Å uncertainty without additional raw data or an explicit, validated interpolation procedure. Because the interpretation of the eight bright dots in Fig. 2c as eight individual C–H bonds and the bond-level Lego construction both rely on this resolution claim, the manuscript needs to present higher-sampled line data or a carefully described and justified resampling/error analysis.","section":"Fig. 1c–e, Results and Discussion"},{"comment":"The structural reconstruction is not de novo: the target molecule, its elemental composition (C, N, H, Mg), and its expected vibrational frequencies are known from the literature before the Lego assembly begins. Each step, from sp2 C–H groups to pyrrole rings, bridging C–H, and the central Mg atom, is anchored in Raman group frequencies and in reported Mg-porphine modes (Refs. 33–38), and the computed SRP images in Supplementary S4 would use the same assumed structure. Consequently, the claim that the chemical structure is 'fully determined in real space' overstates what the demonstration establishes. The authors should either provide a blinded reconstruction, for example by withholding the molecular identity until after the Lego assembly, or explicitly frame the result as a consistency check and a proof of principle that relies on prior knowledge.","section":"Lego assembly, Results and Discussion, Figs. 2–4"},{"comment":"The sign rule for the interference cross-term is asserted rather than derived. The text states that a positive cross-term from symmetric vibrational motion gives constructive signal and a negative cross-term from anti-symmetric motion gives destructive signal, but the derivation of how the normal-mode phases enter the Raman polarizability and survive the integration with g(r − R0) is not given in the main text. This rule is load-bearing for the interpretation of Fig. 2c/d (eight dots vs. four lobes) and for the phase relations among pyrrole rings in Fig. 3b. The authors should provide the full derivation in the main text or a detailed supplementary section, and show that the sign relationship is robust when the confined field covers more than two atomic centers, as the text itself states it applies to multi-center cases.","section":"Eq. (2) and the following paragraph"},{"comment":"The width of the nanocavity field distribution g(r − R0) is a free, unquantified parameter whose value is central to the claimed resolution and to the one-to-one mapping of bright/dark patterns onto bond positions. If the effective optical spot were broader than the 1.5 Å claimed, the cross-term contribution would be smeared and the destructive-interference gaps would not correspond to bond positions. The manuscript should either estimate g(r − R0) from an independent model of the atomistic tip–substrate junction, fit it to the measured line profiles, or report a sensitivity analysis showing how the structural conclusions vary with the assumed field width.","section":"Eqs. (1)–(2), field confinement assumption"}],"minor_comments":[{"comment":"The text refers to the low-frequency mode as '361 cm−1' while the Fig. 4e caption reads '362 cm−1'; please harmonize the values.","section":"Results and Discussion, Fig. 4"},{"comment":"The caption states the scan size and pixel count (2.5 nm × 2.5 nm, 25 × 25 pixels) but not the pixel pitch; please state explicitly that the pitch is 1.0 Å so that the sampling limitation is transparent to the reader.","section":"Fig. 1c caption"},{"comment":"The derivation of Eq. (2), the analysis of the 1475 cm−1 mode, the computed SRP images, and the Supplementary Video are relegated to Supplementary Materials S2–S4, which are not included in the submitted version. To allow verification, the relevant derivations and the quantitative comparison between simulated and experimental images should be made available with the manuscript.","section":"Main text and Supplementary Materials"},{"comment":"For the assignment of the 3072 and 3092 cm−1 modes to sp2 C–H stretching, the general group-frequency reference [33] is cited; a more specific reference or a normal-mode analysis of Mg-porphine would strengthen the assignment.","section":"References, sp2 C–H assignment"},{"comment":"The phrase 'full Raman images of individual vibrational modes on the Ångström level' overstates the demonstrated scope; the paper shows images of selected, observed modes. Consider qualifying this as 'representative individual vibrational modes' or 'all observed Raman modes of the target molecule.'","section":"Abstract and Conclusions"}],"recommendation":"major_revision","confidential_remarks":"The experimental dataset is striking and the SRP concept is likely to attract wide attention, but the resolution claim and the de novo structural determination are the two pillars of the paper, and both need substantial strengthening. Please ensure that the supplementary materials, especially the derivation of the interference sign rule and the computed SRP images, are made available to reviewers; without them the central interpretation cannot be checked."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper shows something real: a set of Raman images for a single Mg-porphine molecule, mode by mode, with features that line up with bond positions. That is a step beyond earlier TERS imaging, including refs 22 and 25, and the Lego-like assembly protocol is a clever way to think about structure determination. If the resolution claim holds, this is a useful new capability.\n\nBut the headline resolution number is shaky. The 25×25 pixel scan over 2.5 nm gives a ~1.0 Å pixel pitch, coarser than half the claimed 1.5 Å FWHM. With a 1.5 Å feature, you get at most two or three independent samples across, so the quoted 1.5(1) Å and its 0.1 Å error cannot be robustly determined from that image alone. The authors need to show either a higher-density scan, the raw line profile before interpolation, or an independent resolution test. Without that, the bond-level interpretation in Figs. 2–4 rests on a number we cannot trust.\n\nThe second soft spot is circularity. The target molecule is known, and the vibrational assignments come from the literature. The Lego build is a reconstruction that re-inserts the expected answer. The computed images in S4 would validate the imaging model, but they cannot validate the ability to determine an unknown structure. A blind or semi-blind test on a molecule not chosen by the same group would materially strengthen the claim. The interference sign rule (symmetric constructive, antisymmetric destructive) is asserted rather than derived in the main text; it is backed by earlier theory from the same group, so it is not a red flag, but it is a place the referee should check.\n\nNone of this undercuts the value of the experimental dataset itself. The mode-by-mode contrast and the apparent correlation with molecular geometry are genuinely interesting. The paper deserves a serious referee, and the concerns are addressable: more sampling, raw data, and a less circular demonstration. I would not block publication on the under-sampling alone, but the authors should be pushed to present the resolution estimate honestly and to soften the claim that this is a general structural determination method at this point.","headline":"Genuine new vibrational-mode imaging for a single molecule, but the 1.5 Å resolution claim is under-sampled and the Lego reconstruction is demonstrated on a known structure; deserves peer review with a blind test.","tokens_in":10189,"tokens_out":2761,"would_cite":true,"duration_ms":27694,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper shows that Ångström-resolved Raman images of individual vibrational modes, read through bond-interference contrast, can determine the complete chemical structure of a single Mg-porphine molecule in real space.","keywords":["scanning Raman picoscopy","tip-enhanced Raman spectroscopy","single-molecule vibrational imaging","Ångström resolution","interference effect","chemical structure determination","Mg-porphine","molecular Lego assembly"],"falsifier":"Deuterating the eight C–H bonds of the molecule should shift the 3072 cm−1 mode to roughly 2200–2300 cm−1; if the eight-dot SRP image does not follow the shifted mode, the bond-level assignment and the interference model are wrong. A second check is to broaden the nanocavity, for example with a blunter tip: the eight-dot pattern should wash out into fewer lobes as the field ceases to sample single bonds.","tokens_in":9158,"feed_emoji":"🔬","tokens_out":5973,"duration_ms":60810,"temperature":0.7,"pith_summary":"Scanning Raman picoscopy is introduced as a way to determine the chemical structure of a single molecule directly in real space. The paper reports full Raman images of individual vibrational modes of one Mg-porphine molecule on Ag(100), with a spatial resolution of 1.5(1) Å, and shows that each mode has its own characteristic spatial pattern. The key step is reading the interference between neighboring bond vibrations: in-phase vibrations add constructively, out-of-phase vibrations cancel, so bright and dark spots mark actual bond positions and connectivity. By overlaying just a few mode images and using Raman group-frequency fingerprints, the authors reconstruct the entire Mg-porphine structure through a Lego-like assembly.","feed_headline":"Raman images at 1.5 Å build a molecule's structure","feed_subtitle":"Full vibrational-mode images let researchers assemble Mg-porphine's connectivity piece by piece from Raman data alone.","key_machinery":"The load-bearing object is the confined nanocavity plasmonic field $g(\\mathbf{r}-\\mathbf{R}_0)$ that appears in Eq. (1), together with the atomic-orbital expansion of the vibronic transition moment in Eq. (2). When the tip is over an atom, the diagonal term dominates; when it is between two atoms, the cross-term between neighboring atomic orbitals becomes important. In-phase (symmetric) vibrations give a positive cross-term and constructive signal; out-of-phase (antisymmetric) vibrations give a negative cross-term and destructive signal. This interference effect turns normal-mode images into a map of bond positions and phase relations, and the Raman fingerprint database supplies the chemical identity of each piece (C–H, pyrrole ring, C–H bridge, Mg–N).","core_discovery":"The paper claims that a single molecule's chemical structure can be fully determined in real space from Ångström-resolved Raman images alone. For a single Mg-porphine molecule, the authors obtain complete spatial maps of many vibrational modes at a resolution of 1.5(1) Å. The 3072 cm−1 mode appears as eight bright dots marking eight sp2 C–H bonds, while the 3092 cm−1 mode has four lobes from constructive interference between neighboring C–H bonds. Pyrrole ring modes in the 1300–1700 cm−1 region show four-lobe patterns whose contrast encodes whether neighboring rings vibrate symmetrically or antisymmetrically, and low-frequency modes locate the central Mg atom via Mg–N vibrations. Combining these images with Raman fingerprint frequencies, the paper shows that the molecule's connectivity can be assembled piece by piece, and computed images agree well with the experimental ones.","pith_inferences":["Beyond the paper: if the sign of the cross-term reliably reports vibrational phase, SRP images could be inverted to recover the full normal-mode eigenvector of a single molecule, not just its structure.","Beyond the paper: the phase-sensitive contrast should be isotope-sensitive, so deuterating specific bonds would provide a direct test and a way to tag chosen groups in a larger molecule.","Beyond the paper: an automated structure-solver that takes a stack of SRP images and returns connectivity would turn the Lego procedure into a general algorithm, which the authors hint at but do not implement.","Beyond the paper: the 1.5 Å resolution and phase readout may also reveal how a molecule deforms when adsorbed on a metal surface, since out-of-plane modes would report the molecule–surface coupling."],"forward_implications":["Chemical structure determination of surface-adsorbed molecules can be done in real space from a handful of Raman images, without crystallization or ensemble averaging.","Full vibrational-mode imaging at 1.5 Å resolution pushes tip-enhanced Raman spectromicroscopy to the single-chemical-bond level.","The Lego-style assembly protocol is intended to generalize to other molecules, and the authors point to machine-learning image recognition as a natural next step.","Combining SRP with noncontact AFM or inelastic tunneling probes should add complementary structural constraints for molecules where Raman fingerprints alone are ambiguous."],"supporting_citations":[{"why":"Supplies the initial chemical mapping of a single molecule by plasmon-enhanced Raman scattering that motivates angstrom-scale SRP.","marker":"[22]"},{"why":"Demonstrates visualization of vibrational normal modes with atomically confined light, the immediate precursor to full-mode imaging.","marker":"[25]"},{"why":"Provides theoretical modeling of subnanometer plasmon-enhanced Raman images used for interpreting mode patterns.","marker":"[30]"},{"why":"Atomistic near-field TERS imaging theory supporting the field-confinement model behind Eqs. (1) and (2).","marker":"[31]"},{"why":"Theory for high-resolution resonant and nonresonant Raman images underlying the calculation of SRP images.","marker":"[32]"},{"why":"Standard Raman group-frequency tables used to assign C–H, ring, and bridge vibrational bands.","marker":"[33]"},{"why":"Scaled quantum-mechanical vibrational spectra of magnesium porphyrins used to assign Mg–N modes.","marker":"[36]"},{"why":"Normal coordinate analysis of metalloporphins supporting the Mg–N vibration frequencies used in the final assembly step.","marker":"[37]"}],"fun_headline_variants":["Ångström Raman shots assemble a molecule atom by atom","Raman picoscopy: see a molecule's bonds, assemble it","Single-molecule Raman images reveal bonds at 1.5 Å","Lego-like Raman images build molecular structure"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument depends on the nanocavity field being narrow enough that the Raman signal at each tip position reports only the molecule directly beneath the tip, with the sign of the cross-term between neighboring atomic orbitals faithfully encoding whether their vibration is in phase or out of phase.","fun_headline_variants_meta":{"raw":{"variants":["Ångström Raman shots assemble a molecule atom by atom","Raman picoscopy: see a molecule's bonds, assemble it","Single-molecule Raman images reveal bonds at 1.5 Å","Lego-like Raman images build molecular structure"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000631,"raw_usage":{"total_tokens":2879,"prompt_tokens":874,"completion_tokens":2005,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":490,"completion_tokens_details":{"reasoning_tokens":1937}},"tokens_in":490,"tokens_out":2005,"duration_ms":16469,"temperature":1.0,"reasoning_tokens":1937,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:31:51.645515+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Deuterating the eight C–H bonds of the molecule should shift the 3072 cm−1 mode to roughly 2200–2300 cm−1; if the eight-dot SRP image does not follow the shifted mode, the bond-level assignment and the interference model are wrong. A second check is to broaden the nanocavity, for example with a blunter tip: the eight-dot pattern should wash out into fewer lobes as the field ceases to sample single bonds.","supporting_citations":[{"cited_title":"Chemical mapping of a single molecule by plasmon - enhanced Raman scattering","cited_arxiv_id":null,"evidence_quote":"Supplies the initial chemical mapping of a single molecule by plasmon-enhanced Raman scattering that motivates angstrom-scale SRP."},{"cited_title":"Visualizing vibrational normal modes of a single molecule with atomically confined light","cited_arxiv_id":null,"evidence_quote":"Demonstrates visualization of vibrational normal modes with atomically confined light, the immediate precursor to full-mode imaging."},{"cited_title":"Theoretical modeling of plasmon - enhanced raman images of a single molecule with subnanometer resolution","cited_arxiv_id":null,"evidence_quote":"Provides theoretical modeling of subnanometer plasmon-enhanced Raman images used for interpreting mode patterns."},{"cited_title":"Single - molecule imaging using atomistic near - field tip - enhanced Raman spectroscop y","cited_arxiv_id":null,"evidence_quote":"Atomistic near-field TERS imaging theory supporting the field-confinement model behind Eqs. (1) and (2)."},{"cited_title":"Theory for modeling of high resolution resonant and nonresonant Raman images","cited_arxiv_id":null,"evidence_quote":"Theory for high-resolution resonant and nonresonant Raman images underlying the calculation of SRP images."},{"cited_title":"Scaled quantum mechanical and experimental vibrational spectra of magnesium and zinc porphyrins","cited_arxiv_id":null,"evidence_quote":"Scaled quantum-mechanical vibrational spectra of magnesium porphyrins used to assign Mg–N modes."},{"cited_title":"Infrared spectra and normal coordinate analysis of metalloporphins","cited_arxiv_id":null,"evidence_quote":"Normal coordinate analysis of metalloporphins supporting the Mg–N vibration frequencies used in the final assembly step."}],"review_version":1}